Derby
  1. Derby
  2. DERBY-4280

Property not documented - derby.language.statementCacheSize

    Details

      Description

      The property derby.language.statementCacheSize is not documented in any of the Derby documentation.

      This property may be useful for tuning Derby.

      In addition, setting this property to 0 is vital for working around DERBY-4279, and its lack of documentation makes it more difficult for users to discover this workaround.

      1. DERBY-4280-3.zip
        6 kB
        Kim Haase
      2. DERBY-4280-3.stat
        0.1 kB
        Kim Haase
      3. DERBY-4280-3.diff
        1 kB
        Kim Haase
      4. DERBY-4280-2.zip
        11 kB
        Kim Haase
      5. DERBY-4280-2.diff
        6 kB
        Kim Haase
      6. DERBY-4280.zip
        11 kB
        Kim Haase
      7. DERBY-4280.stat
        0.2 kB
        Kim Haase
      8. DERBY-4280.diff
        5 kB
        Kim Haase

        Activity

        Hide
        Myrna van Lunteren added a comment -

        modifying list of fix-in versions to see if it then will pop into the 10.10.2 release notes.

        Show
        Myrna van Lunteren added a comment - modifying list of fix-in versions to see if it then will pop into the 10.10.2 release notes.
        Hide
        Kim Haase added a comment -

        Fixes appear in latest alpha manuals.

        Show
        Kim Haase added a comment - Fixes appear in latest alpha manuals.
        Hide
        Kim Haase added a comment -

        Thanks, Knut!

        Committed patch DERBY-4280-3.diff to documentation trunk at revision 1550425.
        Merged to 10.10 doc branch at revision 1550428.
        Merged to 10.9 doc branch at revision 1550431.

        Show
        Kim Haase added a comment - Thanks, Knut! Committed patch DERBY-4280 -3.diff to documentation trunk at revision 1550425. Merged to 10.10 doc branch at revision 1550428. Merged to 10.9 doc branch at revision 1550431.
        Hide
        ASF subversion and git services added a comment -

        Commit 1550431 from Kim Haase in branch 'docs/branches/10.9'
        [ https://svn.apache.org/r1550431 ]

        DERBY-4280 Property not documented - derby.language.statementCacheSize

        Merged DERBY-4280-3.diff to 10.9 doc branch from trunk revision 1550425.

        Show
        ASF subversion and git services added a comment - Commit 1550431 from Kim Haase in branch 'docs/branches/10.9' [ https://svn.apache.org/r1550431 ] DERBY-4280 Property not documented - derby.language.statementCacheSize Merged DERBY-4280 -3.diff to 10.9 doc branch from trunk revision 1550425.
        Hide
        ASF subversion and git services added a comment -

        Commit 1550428 from Kim Haase in branch 'docs/branches/10.10'
        [ https://svn.apache.org/r1550428 ]

        DERBY-4280 Property not documented - derby.language.statementCacheSize

        Merged DERBY-4280-3.diff to 10.10 doc branch from trunk revision 1550425.

        Show
        ASF subversion and git services added a comment - Commit 1550428 from Kim Haase in branch 'docs/branches/10.10' [ https://svn.apache.org/r1550428 ] DERBY-4280 Property not documented - derby.language.statementCacheSize Merged DERBY-4280 -3.diff to 10.10 doc branch from trunk revision 1550425.
        Hide
        ASF subversion and git services added a comment -

        Commit 1550425 from Kim Haase in branch 'docs/trunk'
        [ https://svn.apache.org/r1550425 ]

        DERBY-4280 Property not documented - derby.language.statementCacheSize

        Modified a Reference Manual topic and a Tuning Derby topic.

        Patch: DERBY-4280-3.diff

        Show
        ASF subversion and git services added a comment - Commit 1550425 from Kim Haase in branch 'docs/trunk' [ https://svn.apache.org/r1550425 ] DERBY-4280 Property not documented - derby.language.statementCacheSize Modified a Reference Manual topic and a Tuning Derby topic. Patch: DERBY-4280 -3.diff
        Hide
        Knut Anders Hatlen added a comment -

        +1 to removing those sentences now.

        Show
        Knut Anders Hatlen added a comment - +1 to removing those sentences now.
        Hide
        Kim Haase added a comment -

        Sorry, I meant 10.9. Although DERBY-4279 was fixed in 10.8 as well.

        Show
        Kim Haase added a comment - Sorry, I meant 10.9. Although DERBY-4279 was fixed in 10.8 as well.
        Hide
        Kim Haase added a comment -

        Seems simpler to use this issue than to file a new one. I'm attaching DERBY-4280-3.diff, DERBY-4280-3.stat, and DERBY-4280-3.zip, removing the bug-workaround language from two topics:

        M src/tuning/ctundepth32379.dita
        M src/ref/rrefproperstatementcachesize.dita

        Can someone (Rick?) check and make sure this is the right thing to do? If so, I'll backport it as far as 10.8, where the previous work was done.

        Show
        Kim Haase added a comment - Seems simpler to use this issue than to file a new one. I'm attaching DERBY-4280 -3.diff, DERBY-4280 -3.stat, and DERBY-4280 -3.zip, removing the bug-workaround language from two topics: M src/tuning/ctundepth32379.dita M src/ref/rrefproperstatementcachesize.dita Can someone (Rick?) check and make sure this is the right thing to do? If so, I'll backport it as far as 10.8, where the previous work was done.
        Hide
        Kim Haase added a comment -

        This issue was filed to provide language that works around DERBY-4279, an issue that was fixed some months after we fixed DERBY-4280. But I forgot to file an issue to remove the language from the reference topic and the tuning guide topic. Better late than never ...

        Show
        Kim Haase added a comment - This issue was filed to provide language that works around DERBY-4279 , an issue that was fixed some months after we fixed DERBY-4280 . But I forgot to file an issue to remove the language from the reference topic and the tuning guide topic. Better late than never ...
        Hide
        Kim Haase added a comment -

        Changes have appeared in Latest Alpha Manuals.

        Show
        Kim Haase added a comment - Changes have appeared in Latest Alpha Manuals.
        Hide
        Kim Haase added a comment -

        Thanks again, Rick.

        Committed patch DERBY-4280-2.diff to documentation trunk at revision 1331473.

        Show
        Kim Haase added a comment - Thanks again, Rick. Committed patch DERBY-4280 -2.diff to documentation trunk at revision 1331473.
        Hide
        Rick Hillegas added a comment -

        Thanks Kim. Looks good. +1

        Show
        Rick Hillegas added a comment - Thanks Kim. Looks good. +1
        Hide
        Kim Haase added a comment -

        Thanks again, Rick. I'm attaching DERBY-4280-2.diff and DERBY-4280-2.zip, which I hope make appropriate changes to rrefproperstatementcachesize.dita and ctundepth32379.dita.

        Show
        Kim Haase added a comment - Thanks again, Rick. I'm attaching DERBY-4280 -2.diff and DERBY-4280 -2.zip, which I hope make appropriate changes to rrefproperstatementcachesize.dita and ctundepth32379.dita.
        Hide
        Kim Haase added a comment -

        Thanks, Rick, for all these suggestions. I added a comment to DERBY-4279. I'll file a revised patch for this.

        Show
        Kim Haase added a comment - Thanks, Rick, for all these suggestions. I added a comment to DERBY-4279 . I'll file a revised patch for this.
        Hide
        Rick Hillegas added a comment -

        Hi Kim,

        These changes look good. I agree that the Developer's Guide doesn't have any pre-existing sections which fit this topic. I think your additions to the Reference and Tuning guides are sufficient.

        Some other comments follow...

        rrefproperstatementcachesize:

        statements kept in memory -> prepared statements kept in memory

        I would add more description to the Function section:

        "For more information on the statement cache, see the 'Using the statement cache' subsection of the Derby Tuning Guide.

        This property controls the number of pre-compiled statements which Derby keeps in its statement cache. Consider raising this number if statement preparation is taking too much time."

        If you are going to include the paragraph about setting the property to 0, then I would add a comment to DERBY-4279 so that we will remember to remove that paragraph after we fix that issue.

        ctundepth32379:

        Similar comments: The real point of adjusting derby.language.statementCacheSize is to reduce the amount of time spent preparing statements. The property lets you balance preparation time against the memory consumed by pre-compiled statements. Setting the property to 0 is a hack which works around the bug described by DERBY-4279. When that bug is fixed, we should stop talking about the hack.

        Thanks,
        -Rick

        Show
        Rick Hillegas added a comment - Hi Kim, These changes look good. I agree that the Developer's Guide doesn't have any pre-existing sections which fit this topic. I think your additions to the Reference and Tuning guides are sufficient. Some other comments follow... rrefproperstatementcachesize: statements kept in memory -> prepared statements kept in memory I would add more description to the Function section: "For more information on the statement cache, see the 'Using the statement cache' subsection of the Derby Tuning Guide. This property controls the number of pre-compiled statements which Derby keeps in its statement cache. Consider raising this number if statement preparation is taking too much time." If you are going to include the paragraph about setting the property to 0, then I would add a comment to DERBY-4279 so that we will remember to remove that paragraph after we fix that issue. ctundepth32379: Similar comments: The real point of adjusting derby.language.statementCacheSize is to reduce the amount of time spent preparing statements. The property lets you balance preparation time against the memory consumed by pre-compiled statements. Setting the property to 0 is a hack which works around the bug described by DERBY-4279 . When that bug is fixed, we should stop talking about the hack. Thanks, -Rick
        Hide
        Kim Haase added a comment -

        Attaching DERBY-4280.diff, DERBY-4280.stat, and DERBY-4280.zip, with changes as follows:

        M src/ref/refderby.ditamap
        M src/ref/crefproper22250.dita
        A src/ref/rrefproperstatementcachesize.dita
        M src/tuning/ctundepth32379.dita

        I created a topic for the derby.language.statementCacheSize property and added a bit of info to the Tuning Derby topic on the statement cache. If there are other logical places to put more info, please let me know.

        I found the default value of the property in the Derby code, but I made some not very educated guesses on the scope of the property and whether it is dynamic or static (see the "Derby properties" topic as well as the topic for the property). Probably some of this needs fixing.

        I could not find a logical place in the Developer's Guide to mention this subject; there are several topics on deadlocks, and after that some topics on multi-threaded applications, but none seemed quite right. Suggestions are welcome.

        Please let me know what needs changing or adding.

        Show
        Kim Haase added a comment - Attaching DERBY-4280 .diff, DERBY-4280 .stat, and DERBY-4280 .zip, with changes as follows: M src/ref/refderby.ditamap M src/ref/crefproper22250.dita A src/ref/rrefproperstatementcachesize.dita M src/tuning/ctundepth32379.dita I created a topic for the derby.language.statementCacheSize property and added a bit of info to the Tuning Derby topic on the statement cache. If there are other logical places to put more info, please let me know. I found the default value of the property in the Derby code, but I made some not very educated guesses on the scope of the property and whether it is dynamic or static (see the "Derby properties" topic as well as the topic for the property). Probably some of this needs fixing. I could not find a logical place in the Developer's Guide to mention this subject; there are several topics on deadlocks, and after that some topics on multi-threaded applications, but none seemed quite right. Suggestions are welcome. Please let me know what needs changing or adding.

          People

          • Assignee:
            Kim Haase
            Reporter:
            Jeff Stuckman
          • Votes:
            0 Vote for this issue
            Watchers:
            3 Start watching this issue

            Dates

            • Created:
              Updated:
              Resolved:

              Development